• KIPS Transactions on Computer and Communication Systems
• /
• v.4 no.11
• /
• pp.369-382
• /
• 2015

Energy aware server clusters aim to reduce power consumption at maximum while keeping QoS(quality of service) compared to energy non-aware server clusters. They adjust the power mode of each server in a fixed or variable time interval to activate only the minimum number of servers needed to handle current user requests. Previous studies on energy aware server cluster put efforts to reduce power consumption or heat dissipation, but they do not consider energy efficiency well. In this paper, we propose an energy efficient cluster management method to improve not only performance per watt but also QoS of the existing server power mode control method based on autonomous learning. Our proposed method is to adjust server power mode based on a hybrid approach of autonomous learning method with multi level thresholds and power consumption prediction method. Autonomous learning method with multi level thresholds is applied under normal load situation whereas power consumption prediction method is applied under abnormal load situation. The decision on whether current load is normal or abnormal depends on the ratio of the number of current user requests over the average number of user requests during recent past few minutes. Also, a dynamic shutdown method is additionally applied to shorten the time delay to make servers off. We performed experiments with a cluster of 16 servers using three different kinds of load patterns. The multi-threshold based learning method with prediction and dynamic shutdown shows the best result in terms of normalized QoS and performance per watt (valid responses). For banking load pattern, real load pattern, and virtual load pattern, the numbers of good response per watt in the proposed method increase by 1.66%, 2.9% and 3.84%, respectively, whereas QoS in the proposed method increase by 0.45%, 1.33% and 8.82%, respectively, compared to those in the existing autonomous learning method with single level threshold.

• The KIPS Transactions:PartD
• /
• v.14D no.5
• /
• pp.555-564
• /
• 2007

Internet streaming systems consist of a media server, a streaming sewer, and terminals. The media server delivers multimedia contents such as video and/or audio to the streaming server, which distributes the contents to terminals as well. Existing Internet streaming systems have a bottleneck problem in the streaming server because of the limit of the processing capacity of the streaming server and therefore a streaming server can not accomodate more terminals than the limit. As a solution to this problem, P2P streaming systems have been lately proposed and investigated, using P2P distributed architectures. Actually, however, there exist many difficulties in the design and implementation of P2P streaming systems, because it needs many real computers and various network constructions. In this paper, we have proposed and defined a P2P streaming system model such as the architectural model, the timing model, the behavior model, and the performance metrics. And also we have implemented an NS2 based P2P streaming system simulator called P2PStreamSim. Finally, we have verified it through test simulations and analyzed the results.

• The KIPS Transactions:PartA
• /
• v.14A no.6
• /
• pp.371-382
• /
• 2007

Modern server systems are usually composed in the form of cluster systems in order to serve not only as many users but also as many kinds of applications as possible. The progression of the cluster system architecture leads in a middleware approach based on the Java framework. The middleware approach alleviates the efforts for the construction and the management of a server system but still preserves its performance and applications on the server. In this research, we introduce a new clustering scheme for the easy construction and maintenance of a cluster server system with the Java Management Extensions. We first demonstrate the construction and configuration process. Our experiment sets can verify that it is easy to construct, expand and manage a middleware based cluster system as well as the applications which reside on it. In addition, we can achieve reasonable performance on our service oriented clustered system with the help of state-of-the-art middleware. The experimental results of performance demonstration contain the availability of a server, and the effectiveness of load balancing and scheduling mechanisms. Especially, our service oriented scheduling mechanism was shown to successfully manage load imbalance under the normal load and cope with the overloaded situations, compared with other known scheduling mechanisms.

• The KIPS Transactions:PartC
• /
• v.12C no.4 s.100
• /
• pp.603-608
• /
• 2005

Disk I/O is a major performance bottleneck of web proxy server. Today's most web proxy sowers are design to run on top of a general purpose file system. But general purpose file system can not efficiently handle web cache workload, small files, leading to the performance degradation of entire web proxy servers. In this paper we evaluate the performance potential of raw disk to reduce disk I/O overhead of web proxy servers. To show the performance potential of raw disk, we design a storage management system called Block-structured Storage Management System (BSMS). And we also actually implement web proxy server that incorporate BSMS in Squid. Comprehensive experimental evaluations show that raw disk can be a good solution to improve disk I/O performance significantly for web proxy servers.

• Journal of Computing Science and Engineering
• /
• v.2 no.4
• /
• pp.394-411
• /
• 2008

The data in automation and control systems can be achieved not only from different field devices but also from different OPC (OLE for Process Control) servers. However, current OPC clients can only read and decode the simple data from OPC servers. They will have some problems to acquire structured data and exchange the structured data. In addition to the large network control systems, the OPC clients can read, write, and subscribe to thousands of data points from/to OPC servers. Due to that, the most important factor for building a high performance and scalable industrial control system is the ability to transfer the process data between server and client in the most efficient and fastest way. In order to solve these problems, we propose a means to implement the OPC DA (Data Access) server supporting the OPC complex data, so that the OPC DA clients are able to read and decode any type of data from OPC servers. We also propose a method for caching the process data in large industrial control systems to overcome the limitation of performance of the pure OPC DA system. The performance analysis and discussion indicate that the proposed system has an acceptable performance and is feasible in order for applying to real-time industrial systems today.

• The Journal of the Korea Contents Association
• /
• v.8 no.1
• /
• pp.152-159
• /
• 2008

In this paper, we have studied a new method in which multimedia message can be processed effectively in SIP(Session Initiation Protocol)-based Instant Messaging service. Total system is composed of multimedia message server, presence server and user system. To provide effectively message informations such as number of messages stored in the multimedia message server, waiting status and message types to users, we have proposed a new method which can be a substitute for message waiting indication event package. In our proposed method, when changes of messages status informations occur, these informations are sent to the presence server, and can be transferred as a separate tag within notification message. The presence server processes subscription requests and notifications, and efficiency can be improved as it provides both message status informations and several communication status informations such as on-line status to service users. Partial publication is used in our system to improve data transfer efficiency more between user system and the presence server. The performance of our proposed model is evaluated by experiments measuring multimedia message transfer speed and server processing time.

• The KIPS Transactions:PartC
• /
• v.16C no.5
• /
• pp.597-604
• /
• 2009

An SMTP(Simple Mail Transfer Protocol) server was designed for delivering and storing user's email across a network. An SMTP server can be distributed as multiple servers for large service, huge users or massive emails. An SMTP server can be constructed by multiple servers with separating file system as email storing space, and each SMTP server can usually share each file system by using the NFS(Network File system). However the NFS is originally designed for sharing each file system across a network, and contains all attributes and features of regular file system. Using this NFS for email storing space of SMTP servers, it makes overhead due to unnecessary work of regular file system. In this paper, we propose a method to do directly operation with socket for reducing work overhead caused by the NFS. For doing directly operation with socket, this paper defines information based storing space structure and a protocol for storing emails. We performed experiments using Mailstone made by Netscape. The experimental results show the performance improvement of the proposed method compared to the existing method.

• Journal of IKEEE
• /
• v.5 no.2 s.9
• /
• pp.174-181
• /
• 2001

In this study, the architecture of the Web Camera System for realtime monitoring on Internet is proposed and implemented in two different structures. In the one architecture, a Web-server and a Camera-server are implemented on the same system, and the system transfers motion pictures compressed to JPEG file to users on the WWW(World Wide Web). In the other architecture, the Web-server and the Camera-server are implemented on different systems, and the motion pictures are transferred from the Camera-server to the Web-server, and finally to users. For JPEG image transferring in the Web Camera system, the Java Applet and the Java Script are used to maximize flexibility of the system from the Operating system and the Web browsers. In order to compare system performance between two architectures, data traffic is measured and simulated in the unit of byte per second.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.41 no.4
• /
• pp.67-72
• /
• 2004

This paper proposes a method improving the convergence speed of optimal solution for parallel genetic algorithm in the network based client-server model. Unlike the existing methods of obtaining global elite only by evaluating local elites in server, the proposed method obtains it by evaluating local elites and improving its fitness by applying genetic algorithm during idle time of the server. By using the improved chromosome in server for the client's genetic algorithm processing, the convergence speed of the optimal solution is increased. The improvement of fitness at the server during the interval of chromosome migration is (equation omitted)(F$_{max}$(g)-F$_{max}$(g-1)), whole F$_{max}$(g) is a max fitness of the g-th generation and G is the number of improved generation by the server. As the number of clients increases and G decreases, the improvement of fitness goes down. However the improvement of fitness is better than existing methods..

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.13 no.4
• /
• pp.29-37
• /
• 2013

Multiplayer games typically organized based on a client-server(CS) or peer-to-peer(PP) architecture. The CS architecture is not scalable with the number of players due to a large bandwidth requirement at the server. The PP architecture, on the other hand, introduces significant overhead for the players, as each player needs to check the consistency between its local state and the state of all other players. We then propose a method that combines the merits of CS and PP. In this method, players exchange updates with lower priority in a peer-to-peer manner but communicate directly with a central server for the other updates. As a result, the proposed method has a lower network bandwidth requirement than the server of a CS architecture and the server bandwidth bottleneck is removed. For another important issue about multiplayer games, this method always maintains state consistency among players correctly. The performance of this method is evaluated through extensive simulation experiments and analysis.